AFAIK my DAC only uses the PCM2704 for the USB interface, which I am not using.
I can't get a schematic, but looking at the pcb the coaxial signal goes to the DIR9001, which is 24bit/96KHz capable.
The signal I'm feeding comes from my computer via network to my Mede8er media player, from there coaxial to the DAC.
I can't get a schematic, but looking at the pcb the coaxial signal goes to the DIR9001, which is 24bit/96KHz capable.
The signal I'm feeding comes from my computer via network to my Mede8er media player, from there coaxial to the DAC.
Last edited:
The LM6172 is incredibly FAST at 3,000V/uSec, but its distortion figures (~-103db) and noise figures (12nV√Hz voltage and 1pa√Hz current) are not really competitive with many other opamps that could be used here (LM4562,OPA1612,OPA2227,OPA1642,AD8599). In addition, it has woefully poor CMRR(-70db).
Maybe because lowest possible noise and lowest possible distortion is what we are after ? 😉 I mean using the best specced parts won't hurt.
Last edited:
Why would you be after those aims? As opposed to low enough distortion and low enough noise for the particular application?
So back the original question - why would lowest possible noise and distortion be appropriate to this particular application?
To get the best SOUND!! But, in the area of "good enough", that really depends on your pocketbook. For instance, I won't consider OPA627s because they are $25 apiece from Mouser (my prefered distributor)---way TOO much $$$ for my projects. My favorites are the LM4562 (~$1.50 per opamp); a bi-polar for the low-to-medium impedance circuits; the OPA1642 FET (also ~$1.50 per opamp) for the high impedance applications (guitar amp). Both are inexpensive, dual-8-packaged (although the 1642 does require an SOIC-to-DIP adaptor), so they're convenient and fit MY budget!!
In this particular application (DAC output), the noise spec of any if the good opamps is way below audibility. Noise figures only really matter when in a high-gain situation--preamplifiers for microphone, phono cartridge, or guitar. For line level, I try to get the lowest distortion for the buck. And again, the sound differences between the premium opamps I mentioned before (LM4562,OPA1612,OPA2227,OPA1642,AD8599) are really minute and probably undetectable in a double-blind test.
Last edited:
Right on there! But there's a pretty tenuous link between the numbers you're seeking in distortion and noise and getting the best sound. I rate the LM6172 for sound in this application despite its numbers being inferior in some respects to other parts.
Below audibility and considerably below the noise of the vast majority of recordings in that circuit position. Which makes noise a non-issue here.
I have been reading lots of DAC tests, going from the very low noise LT1028 and AD797, and lower noise seems not to be something that makes them sound better in that application.
Maybe because low noise chips generally have a slower slew rate, and SR does seem to have a more important relation to sound quality.
Lower distortion chips also are not on the "chosen" list, at least not in present day DACs, which are all voltage output types, not current types.
Large BW, high slew rate (> 50v/us) and medium noise chips seem to be a better bet. But in the end it seems to have much to do with personal preferences. Very much like transistors and tubes. That's how it looks to me.
Maybe because low noise chips generally have a slower slew rate, and SR does seem to have a more important relation to sound quality.
Lower distortion chips also are not on the "chosen" list, at least not in present day DACs, which are all voltage output types, not current types.
Large BW, high slew rate (> 50v/us) and medium noise chips seem to be a better bet. But in the end it seems to have much to do with personal preferences. Very much like transistors and tubes. That's how it looks to me.
Last edited:
The application of DAC I/V is particularly onerous for an opamp as the slew rates out of a CMOS-process DAC chip easily run into volts/nS. A low noise bipolar opamp can't afford input degeneration resistors (they'll add too much noise) so this means the input LTP will have an extremely narrow relatively linear region, which the DAC's impulsive output will violate at each update. So there's an argument for better linearity in the input stage - one way to achieve that is LTP degeneration (at the expense of noise). A part which features input degeneration is the AD847 (AD827 is the dual version).
A faster slew rate may mean the opamp settles sooner - meaning it spends less time at the wrong value, though there's a lot more to settling time than slew rate.
I/V conversion is non-existent if we are talking about voltage output DAC types, like the AK4396 or some others. The question seems just LP filtering and buffering.
The AD825 is a fast jFet option, with reasonably low noise, but not everybody likes it.
In my case I have two DAC boards that I want to improve on the outputs: one with AD1955 and another with a not assembled yet AK4396.
The first one came with with an instrumentation 3-chip output using OP275s. I am trying chips on it. Now it's using AD826, and the sound is not much different.
I have LM6172s, but I'm not sure if I don't need some tweaking.
The AD825 is a fast jFet option, with reasonably low noise, but not everybody likes it.
In my case I have two DAC boards that I want to improve on the outputs: one with AD1955 and another with a not assembled yet AK4396.
The first one came with with an instrumentation 3-chip output using OP275s. I am trying chips on it. Now it's using AD826, and the sound is not much different.
I have LM6172s, but I'm not sure if I don't need some tweaking.
Wow, so there's a pretty likely chance that my opa627s are something else!
I have access to a high powered camera that could tell me, now that Mr. Wurcer has a picture of the real thing.
They sounded better than any of the other op amps I had used before however, and passed the ohms test that gauged r between pins 1,5 if I recall, whatever that was worth...
Whatever op amps you wind up with, good power decoupling will nearly always make a difference.
I have access to a high powered camera that could tell me, now that Mr. Wurcer has a picture of the real thing.
They sounded better than any of the other op amps I had used before however, and passed the ohms test that gauged r between pins 1,5 if I recall, whatever that was worth...
Whatever op amps you wind up with, good power decoupling will nearly always make a difference.
- Status
- Not open for further replies.